Understanding Game Behaviour Among Adolescents

Team cuRiosity

How do players behave as they progress through the game?

We present the results of our exploratory research that aims to:

  • characterise play style by:
    • investigating variations between different demographics
    • tracking changes in skill prioritization over the progression of the game
    • understanding the perseverance of players on their number of attempts
  • player perception vs game progression

3 Ps:

  • Prioritisation

  • Perseverance

  • Perception

Background on the data used:

  • This data is from a video game called My Future is My Life (Code named: Ivy). This data compiles participant’s data and uses it to reduce risk taking behaviours and lowering the spread of human immunodeficiency virus (HIV) amongst adolescents.

Data used

  • Demographic information of the created avatars, i.e. age, gender, ethnicity (this is for avatars only)

  • Happiness rating how satisfied the players were from their epilogue endings

  • Skill level avatar’s final skill level at the end of the game

  • Risk metric in a playthrough each bad decision leads to a strike and after 3 strikes you fail that scenario

  • Percentage complete progression in the game to 100% i.e. player starts in a bad state and ends with an optimal state

Category 1: Prioritisation

How did players prioritise different skills?

Avatar Gender

Avatar Ethnicity

Risk Metric

How did players progress through their skills?

Avatar Ethnicity

Avatar Gender

Schools

Category 2: Perseverance

How did players prioritise different skills?

Player attempts

By Gender

By Ethnicity

By School

Category 3: Perception

How did players prioritise different skills?


Call:
lm(data = plotted_data)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.19093 -0.02680  0.01298  0.03892  0.19442 

Coefficients:
                     Estimate Std. Error t value Pr(>|t|)    
(Intercept)           0.59776    0.04097   14.59 2.87e-10 ***
percentage_completed -0.87059    0.08612  -10.11 4.33e-08 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 0.09996 on 15 degrees of freedom
Multiple R-squared:  0.872, Adjusted R-squared:  0.8635 
F-statistic: 102.2 on 1 and 15 DF,  p-value: 4.33e-08

Ethnicity

Gender

Risk Levels

Future Recommendations

  • Analyse survey results with the different play styles of players
  • Determine whether outlets like this game can be used as a suitable method of teaching adolescents about risk behaviours
  • What extent can these outlets can help in reducing the risk taking behaviour evident in the bad decisions and endings of the games?

Conclusion

  1. Effect of demographics was consistent across prioritisation and perseverance - no significant effect
  2. There is a strong potential for using the characteristics of perception as significant predictor of identifying individuals who are more inclined towards risk taking behaviours.